A plot of inspiral range hour-trends supports the impression that the rate of major glitches is, at present, greater than in O1 or O2 at LHO and possibly LLO as well. The minimum range per hour in Figure 1 (blue) appears lower now than in O1 and O2 (and the associated commissioning time) because fewer hours are free of major range drops. The plot on the second page of Figure 1 compares single recent days around the supernova alert, when the interferometers were mostly left free of commissioning glitches, to random days in O2, and these plots also suggest a higher big glitch rate at both sites. Hopefully, the high glitch rate will go away once the interferometers are in run mode, but I think we should start worrying, just in case.

One possibility is that the apparent increase in glitch rate is due to a new piece of equipment such as electronic or computer equipment, that was installed at each site during the O1-O2 break. Sheila pointed out that the possibilities might be narrowed down at LLO since the installation was staggered. I looked for a step increase, and, while it is not completely clear,  the best candidate appeared to be the March – September 2018 down-time (Figure 2). Suggestions of equipment to investigate would be appreciated.

I, and as far as I know, others, have not found channels that predict the majority of the major glitches. So, Kara and I set up magnetometer channels to monitor new equipment that might be too far from the permanent magnetometers for glitch source identification.

Robert, Kara